seasharp
/
anvil
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

* Updated anvil-provision-server to handle startup when the peer doesn't create/connect it's DRBD resource (ie: node is offline).

Browse Source 
Signed-off-by: digimer <digimer@gravitar.alteeve.com>
main
digimer 2 years ago
parent a5cee52153
commit b666caec64
4 changed files with 123 additions and 18 deletions
Whitespace
Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL
Split View
Diff Options
Show Stats Download Patch File Download Diff File
  1. 3
      notes
  2. 4
      ocf/alteeve/server
  3. 13
      share/words.xml
  4. 121
      tools/anvil-provision-server

3
notes
Unescape View File

@ -1,4 +1,7 @@
Add 'lsof' to Required
When pairing Striker, make sure new config goes to all known nodes!
Immediately set drbdadm to 'secondary' after 'primary --force'

4
ocf/alteeve/server
Unescape View File

@ -199,8 +199,8 @@ $anvil->Log->entry({source => $THIS_FILE, line => __LINE__, 'print' => 0, level
foreach my $key (sort {$a cmp $b} keys %{$anvil->data->{environment}})
{
$anvil->Log->variables({source => $THIS_FILE, line => __LINE__, level => 2, list => {
"environment::${key}" => $anvil->data->{environment}{$key},
$anvil->Log->variables({source => $THIS_FILE, line => __LINE__, level => 2, list => {
"environment::${key}" => $anvil->data->{environment}{$key},
}});
}
foreach my $key (sort {$a cmp $b} keys %{$anvil->data->{switches}})

13
share/words.xml
Unescape View File

@ -571,7 +571,8 @@ The definition data passed in was:
<key name="error_0396">[ Error ] - The Anvil! UUID: [#!variable!uuid!#] was not found.</key>
<key name="error_0397">[ Error ] - The DR link UUID: [#!variable!uuid!#] was not found.</key>
<key name="error_0398">[ Error ] - There was a problem processing the requested network: [#!variable!network!#]. Details should be logged.</key>
<key name="error_0399">[ Error ] - It looks like the new device: [#!variable!resource!#] failed to appear. Unable to proceed.</key>
<!-- Files templates -->
<!-- NOTE: Translating these files requires an understanding of which lines are translatable -->
<!-- comments and which are command lines that can't be changed! -->
@ -1227,7 +1228,7 @@ Use the 'Short ID' that best matches your OS.</key>
<key name="job_0215">The server has been flagged as deleted now.</key>
<key name="job_0216">The server delete is complete on this host!</key>
<key name="job_0217">It looks like ScanCore has not yet run on one or both nodes in this Anvil! system. Missing resource data, so unable to proceed.</key>
<key name="job_0218">Manually calling 'scan-drbd' to ensure that the new agent is recorded.</key>
<key name="job_0218">Manually calling 'scan-drbd' to ensure that the new resource is recorded.</key>
<key name="job_0219">The server name: [#!variable!server_name!#] is already used by another server.</key>
<key name="job_0220">Deleting the server's definition file: [#!variable!file!#]...</key>
<key name="job_0221">The server: [#!variable!server_name!#] was not found in the cluster configuration. This can happen if a server was partially deleted and we're trying again.</key>
@ -1476,7 +1477,13 @@ Note: This is a permanent action! If you protect this server again later, a full
<key name="job_0428">The server: [#!variable!server!#] is still running two minutes after asking it to stop. It might have woken up on the first press and ignored the shutdown request (Hi Windows). Pressing the poewr button again.</key>
<key name="job_0429">Copying the Long-throw (drbd proxy) license file: [#!variable!file!#] into place.</key>
<key name="job_0430">The fence device: [#!variable!device!#] no longer has a port associated with it, will remove it.</key>
<key name="job_0431">Calling drbdadm adjust to load the new resource, then waiting for the DRBD device to appear.</key>
<key name="job_0432">The new DRBD resource: [#!variable!resource!#] now exists!</key>
<key name="job_0433">Still waiting for the new DRBD resource: [#!variable!resource!#] to appear...</key>
<key name="job_0434">Waiting for up to a minute to see if the peer connects before provisioning the server.</key>
<key name="job_0435">One or more peer disk state or roles are 'unknown', waiting: [#!variable!waiting!#] seconds longer.</key>
<key name="job_0436">Peer disk state or role still unknown after one minute, proceeding without it.</key>
<!-- Log entries -->
<key name="log_0001">Starting: [#!variable!program!#].</key>
<key name="log_0002">

121
tools/anvil-provision-server
Unescape View File

@ -705,7 +705,17 @@ sub provision_server
output => $output,
return_code => $return_code,
}});
# Call an adjust in case we forced the resource to disable fencing.
$shell_call = $anvil->data->{path}{exe}{drbdadm}." adjust ".$anvil->data->{job}{server_name};
$anvil->Log->variables({source => $THIS_FILE, line => __LINE__, level => 2, list => { shell_call => $shell_call }});
($output, $return_code) = $anvil->System->call({shell_call => $shell_call});
$anvil->Log->variables({source => $THIS_FILE, line => __LINE__, level => 2, list => {
output => $output,
return_code => $return_code,
}});
return(0);
}
@ -758,8 +768,8 @@ sub startup_resource
task => "up",
});
# If both sides are Inconsistent, for node 1 to primary
# If both sides are Inconsistent, for node 1 to primary. If we're primary and the peer is
# Unknown, force to primary.
my $waiting = 1;
while($waiting)
{
@ -924,8 +934,93 @@ sub startup_resource
$anvil->Log->variables({source => $THIS_FILE, line => __LINE__, level => 2, list => { all_ready => $all_ready }});
if ($all_ready)
{
# If we're here, we're mostly ready, however we're in a bit of a spot if our
# peer hasn't connected. We're going to wait up to one minute for the peer to
# connect. If it doesn't, we're going to force back to primary.
$anvil->Job->update_progress({
progress => 59,
message => "job_0434",
});
$anvil->Log->entry({source => $THIS_FILE, line => __LINE__, 'print' => 1, level => 1, key => "job_0434"});
my $wait_until = time + 60;
my $resource = $anvil->data->{job}{server_name};
$anvil->Log->variables({source => $THIS_FILE, line => __LINE__, level => 2, list => {
's1:time' => time,
's2:wait_until' => $wait_until,
's3:resource' => $resource,
}});
while ($waiting)
{
$anvil->DRBD->gather_data({debug => 3});
my $any_unknown = 0;
foreach my $volume (sort {$a cmp $b} keys %{$anvil->data->{new}{resource}{$resource}{volume}})
{
$anvil->Log->variables({source => $THIS_FILE, line => __LINE__, level => 2, list => { volume => $volume }});
foreach my $peer (sort {$a cmp $b} keys %{$anvil->data->{new}{resource}{$resource}{volume}{$volume}{peer}})
{
$anvil->Log->variables({source => $THIS_FILE, line => __LINE__, level => 2, list => { peer => $peer }});
if (not defined $anvil->data->{new}{resource}{$resource}{volume}{$volume}{peer}{$peer}{local_disk_state})
{
$any_unknown = 1;
$anvil->Log->variables({source => $THIS_FILE, line => __LINE__, level => 2, list => { volume => $volume }});
next;
}
my $local_disk_state = $anvil->data->{new}{resource}{$resource}{volume}{$volume}{peer}{$peer}{local_disk_state};
my $peer_disk_state = $anvil->data->{new}{resource}{$resource}{volume}{$volume}{peer}{$peer}{peer_disk_state};
my $local_role = $anvil->data->{new}{resource}{$resource}{volume}{$volume}{peer}{$peer}{local_role};
my $peer_role = $anvil->data->{new}{resource}{$resource}{volume}{$volume}{peer}{$peer}{peer_role};
$anvil->Log->variables({source => $THIS_FILE, line => __LINE__, level => 2, list => {
's1:local_disk_state' => $local_disk_state,
's2:peer_disk_state' => $peer_disk_state,
's3:local_role' => $local_role,
's4:peer_role' => $peer_role,
}});
if (($peer_disk_state =~ /unknown/i) or ($peer_role =~ /unknown/i))
{
$any_unknown = 1;
$anvil->Log->variables({source => $THIS_FILE, line => __LINE__, level => 2, list => { volume => $volume }});
}
last if $any_unknown;
}
last if $any_unknown;
}
if ($any_unknown)
{
if (time > $wait_until)
{
$anvil->Log->entry({source => $THIS_FILE, line => __LINE__, 'print' => 1, level => 1, key => "job_0436"});
my $shell_call = $anvil->data->{path}{exe}{drbdsetup}." net-options ".$anvil->data->{job}{server_name}." ".$anvil->data->{job}{drbd_peer_node_id}." --set-defaults --_name=".$anvil->data->{job}{peer_short_name}." --protocol=C --fencing=dont-care";
$anvil->Log->variables({source => $THIS_FILE, line => __LINE__, level => 2, list => { shell_call => $shell_call }});
my ($output, $return_code) = $anvil->System->call({shell_call => $shell_call});
$anvil->Log->variables({source => $THIS_FILE, line => __LINE__, level => 2, list => {
output => $output,
return_code => $return_code,
}});
$waiting = 0;
$anvil->Log->variables({source => $THIS_FILE, line => __LINE__, level => 2, list => { waiting => $waiting }});
}
else
{
my $waiting = $wait_until - time;
$anvil->Log->entry({source => $THIS_FILE, line => __LINE__, 'print' => 1, level => 1, key => "job_0435", variables => { waiting => $waiting }});
sleep 5;
}
}
else
{
$waiting = 0;
$anvil->Log->variables({source => $THIS_FILE, line => __LINE__, level => 2, list => { waiting => $waiting }});
}
}
$waiting = 0;
$anvil->Log->variables({source => $THIS_FILE, line => __LINE__, level => 2, list => { waiting => $waiting }});
}
if ($waiting)
@ -953,13 +1048,13 @@ sub create_md
# Create the DRBD metadata
my $shell_call = $anvil->data->{path}{exe}{drbdadm}." --force create-md --max-peers=3 ".$anvil->data->{job}{server_name};
$anvil->Log->variables({source => $THIS_FILE, line => __LINE__, level => 2, list => { shell_call => $shell_call }});
my ($output, $return_code) = $anvil->System->call({shell_call => $shell_call});
$anvil->Log->variables({source => $THIS_FILE, line => __LINE__, level => 2, list => {
$anvil->Log->variables({source => $THIS_FILE, line => __LINE__, level => 2, list => {
output => $output,
return_code => $return_code,
}});
### Return codes
# 0 == Success
# 1 == ?
@ -969,21 +1064,21 @@ sub create_md
# Metadata creation failed.
$anvil->Job->update_progress({
progress => 100,
message => "error_0204,!!return_code!".$return_code."!!,!!output!".$output."!!",
job_status => "failed",
message => "error_0204,!!return_code!".$return_code."!!,!!output!".$output."!!",
job_status => "failed",
});
$anvil->Log->entry({source => $THIS_FILE, line => __LINE__, 'print' => 1, level => 0, priority => 'err', key => "error_0204", variables => {
$anvil->Log->entry({source => $THIS_FILE, line => __LINE__, 'print' => 1, level => 0, priority => 'err', key => "error_0204", variables => {
return_code => $return_code,
output => $output,
output => $output,
}});
$anvil->nice_exit({exit_code => 1});
}
$anvil->Log->entry({source => $THIS_FILE, line => __LINE__, 'print' => 1, level => 1, key => "log_0579", variables => { resource => $anvil->data->{job}{server_name} }});
# If we're not the peer, force this resouroce to Primary.
if (not $anvil->data->{job}{peer_mode})
{
}
$anvil->Job->update_progress({

Write Preview
Loading…
Cancel
Save